Havas bets on AI veteran Sharona Sankar-King to lead proprietary tech push
Havas Media Network North America has named Sharona Sankar-King chief data and product officer, giving her the keys to Converged.AI and the agency's analytics practice across the U.S. and Canada. She reports to North America CEO Greg James. The mandate: turn AI, data, and product into one cohesive engine that moves client outcomes.
Sankar-King brings 25+ years across Bain & Company, BBDO, GroupM, and Harte Hanks. That mix of consulting rigor, agency execution, and data science is exactly what Havas says it needs as AI, data, and product disciplines compress into a single operating model.
Why this matters for product development
Havas isn't chasing shiny features. It's building infrastructure and workflows that scale across 23,000 people and dozens of client environments. That's the difference between a demo and a dependable platform.
- Platform > projects: Converged.AI sits beneath tools like AVA, a no-code app that lets teams create and share AI workflows. Product teams should think in platforms, not point solutions.
- Data unity before model novelty: Sankar-King called out the real blockers-fragmented data, disconnected workflows, weak infrastructure, and confusion around AI's role. Fix those first; models come second.
- AI as augmentation: The goal isn't to replace teams; it's to improve precision, speed, and decision quality. Build for human-in-the-loop by default.
- Scale with guardrails: Governance, lineage, and monitoring are not optional. Use frameworks like the NIST AI RMF to set policy and measurement from day one.
Inside Converged.AI and AVA
Converged.AI is the backbone of Havas' AI stack, connecting teams across the global network. It underpins shared tools, data services, and repeatable workflows-so each team doesn't rebuild the same thing ten different ways.
On top sits AVA, a no-code builder launched at CES that lets staff spin up and distribute AI workflows. For product orgs, this signals a shift from isolated experimentation to reusable components, templates, and internal marketplaces.
A clear stance on AI's real work
Greg James said the industry has shifted fast, and Havas needed a leader who could unify AI strategy, product, and client impact. That's where Sankar-King steps in.
Her focus: "accelerate our ability to deliver intelligence-led solutions at scale." She called out a common trap-surface-level automation and siloed pilots that ignore the plumbing required for real value.
Product takeaways you can apply now
- Make AI a product, not a project: Define a platform charter, shared services, and internal APIs. Ship modules; retire one-off experiments.
- Data contracts > ad hoc extracts: Standardize schemas, governance, and lineage so models don't crumble under edge cases and compliance asks.
- Workflow is the UX: Map how people actually work. Build no-code/low-code paths for common tasks; reserve advanced tools for specialists.
- Measure throughput and quality: Track cycle time, accuracy, cost-to-serve, and model adoption. Tie every feature to a business KPI.
- Privacy by design: Bake consent, regional rules, and audit trails into the platform. Don't bolt them on.
- Operate AI like a product line: Stand up LLMOps/MLOps, incident response, versioning, and release notes. Treat prompts and policies as code.
What to watch next
- How Converged.AI standardizes data products and shared workflows across markets.
- Expansion of AVA templates for media, analytics, and creative ops-and how non-technical teams adopt them.
- Clear ROI stories: reduced cycle times, improved targeting accuracy, and lower cost-to-serve across accounts.
- Governance maturity: model registries, evaluation pipelines, bias testing, and auditability at scale.
If you're building internal AI platforms, this move is a nudge: invest in the plumbing, create shared components, and ship outcomes-quietly and consistently. That's how AI becomes part of how the business works, not a string of demos.
Want structured guidance for your team? Explore AI for Product Development for strategy, prototyping, and platform playbooks.
Your membership also unlocks:
